Action Recognition for Special Needs Students: An Algorithmic Study Integrating Convolutional Neural Networks and Temporal Attention Mechanism
Fan Zhang · 2024
In recent years, due to the continuous development of computer vision, artificial intelligence and other disciplines, higher requirements have been put forward for the training of special students. The current physical education teaching model is difficult to combine with the personal characteristics of special students, especially the accuracy and real-time feedback of the movement characteristics of special students are poor. In response to this issue, this article applied an algorithm that Integrated Convolutional Neural Network and Temporal Attention Mechanism (ICNN-TAM) to accurately identify the motion features of special needs students, and provided personalized guidance and training for them. Research has found that through algorithmic-assisted teaching and training, the accuracy of special needs students' movements is improved, with a maximum recognition accuracy of 94.7%, and can provide teachers with more efficient teaching methods. This system promotes personalized and refined teaching for special needs students' movement training and guidance.